Wikidata talk:WikiProject Epigraphy

Standards for items of epigraphs in Latin

Hi. As part of Wikidata:Testi_latini I will start to create 100s items of epigraphs in Latin from Italy, from Wikimedia Commons and other databases. As stated also on the main page, I have decided to create this project to coordinate the quality of these items in the future. Everything started from these tests

Let's all agree on a possible standard. For example, let's consider CIL-related items. Of course, I use items with CIL since it is a database whose items are more structured already on Commons, but almost all the other metadata could be present in other items of epigraphs.

1. Do you agree with main form of description and aliases of CIL here? (note: it's from the Commons category)
2. Do you agree with the use of collection (P195) for example here?
3. Do you agree with the use of location of discovery (P189) for example here?
4. Do you agree with the use of writing system (P282) and Latin script (Q8229) for example here?
5. Do you agree with the use of width (P2049), height (P2048) and horizontal depth (P5524) for example here?
6. Do you agree with the use of country (P17) for example here?
7. Do you agree with the use of instance of (P31) and archaeological find (Q10855061) for example here or archaeological artifact (Q220659) here?
8. Do you agree with the use of creator (P170) and unknown value for example here?

If so, please help me collect a sufficient amount of examples to be put on the main page so I can follow such guidelines as much as possible.

For sure, I need to pinpoint some core metadata on which we can all agree, this way I or someone else don't have to refine them again soon. Additional details can be added later. --Alexmar983 (talk) 20:27, 13 June 2022 (UTC)

They all seem reasonable; anyway, I suggest to substitute normal bullets with numbered bullets, so as to make it easier to refer to single points and to add language of work or name (P407) in the point 4. --Epìdosis 14:34, 14 June 2022 (UTC)

Done. language of work or name (P407) was IMHO implicit. A written text has by default a language. I will add in the final scheme of a "good item" for the first guideline of this project,--Alexmar983 (talk) 15:07, 14 June 2022 (UTC)

Thanks for launching the project! There is a lot to unpack before we can mass produce Qid of inscriptions, but this is an opportunity to shape this project in the good direction with an agreed standard. Some quick considerations, as I am busy with exams and don't have time to develop:

1) We are missing aliases without the leading 0s, they are the one that would be used in a normal scientific publication.

2) I'm conflicted. I think published in (P1433) should be used as the main indicator to track the books in which they are referenced, though obviously the publication is only a latter step in the object's life. Corpora tend to be published several time and the numerotation change, like in the various editions of the Inscriptiones Graecae (Q673341) or the Sylloge of Dittenberg. And not all inscriptions are published in neat collections.

3) No, I think it should be the modern location, not the ancient one.

4) Yes.

5) See 9)

6) Yes but see 9)

7) See 9)

8) Yes, though it will be so on most items.

9) So one of the biggest issue is: do we consider the inscription as object + text or do we need two items (like we do for books)? I'd prefer the first solution, with some exceptions, mainly for thoses inscriptions that reproduce the same text, e. g. the Res Gestae Divi Augusti (Q734095). But, this means that a Qid must be instance of (P31)inscription (Q1640824) and something else, as inscription (Q1640824) superclasses seems to be too generic (media (Q340169)) on the object side. Same goes for archaeological artifact (Q220659) and archaeological find (Q10855061) IMO: they are just too wide to be of any value by themselves. We must therefore add a P31 to say if its a stela, an altar, a sarcophagus, etc.

I wonder if it could be a good idea to contact Europeana Network for Ancient Greek and Latin Epigraphy (Q18563758) to see if they already have done some thought on it. I know they use Wikibase on their project, so surely they are points of contact. --Jahl de Vautban (talk) 18:20, 14 June 2022 (UTC)

Jahl de Vautban on Friday I will meet with User:Epìdosis and I will show him a summary of the comments here before publishing a small standard for the approval of the project. Basically, I need to create some decent items that can be enlarged later, and not changed related to their core aspects. I can say, that we have few points on which we can for sure agree.

IMHO there is only a major aspect, which I also wrote in some preliminary comments years ago myself, and that you point out here... There are various statements that we can agree are appropriate, but the instance remains the most critical one.

Since I am entering this domain from the perspective of cultural heritage and not wikisource, and considering the inscription is an additional refinement of the object that contains it, I would probably consider as a priority to create items related primarly to physical objects. The items related to text look more an additional refinement of this process. It's a process that will take few years to get to a good in-depth coverage. Still, the object exist before its content, so we have to start logically from the item of the object. However, The CIL is related to inscription, so this is something I have to think carefully.--Alexmar983 (talk) 20:30, 14 June 2022 (UTC)

@Jahl de Vautban: Thanks for your comments! I effectively agree on 3); for 1) could you make an example?; for 2), I tend to think that you are right preferring published in (P1433) to collection (P195) - and the value should probably be an edition of the work instead of the work itself, do you agree? Thus we would need some edition items for CIL editions etc.; regarding 9), it's of course the biggest issue and I would agree with your solution, separating text and material object only in the rare cases when the text is present in more objects. If you are able to contact EAGLE, which in the past had also some collaboration with Wikidata regarding its thesaurus, it would surely be interesting as modeling issues are always the hardest to be solved and experts' opinions are really useful (and I have never attended a specific course on epigraphy in my classical studies, in fact ...). --Epìdosis 06:51, 15 June 2022 (UTC) P.S. I think that the standard we are establishing should in fact not be limited to Latin inscriptions, but regard also the ones in other languages and scripts, which will just have different values on 4)

@Epìdosis: for 1), it is fairly simple : CIL III 4591, CIL 3, 4591, etc. (like CIL VIII, 12908 (Q106618586) in fact). The leading 0s are used on databases for technicals reasons (Clauss-Slaby coming in my mind first), but in a published text you would never find the inscription cited with them. For 2) I did some more reflexion on it: first, to answer Alexmar983, my stance didn't come from my Wikisource background actually, more on my use of inscriptions when I was still doing some courses and studies with them. But I think textual and physical points of view aren't incompatible: in fact, what if we limit the use of collection (P195) to the institution holding the physical object? I think it would make more sense and be a better fit for the property. Then for Epìdosis, yes, published in (P1433) used with the correct edition (even more like the correct volume). I'll try to contact EAGLE, but it will not be before July. --Jahl de Vautban (talk) 15:18, 15 June 2022 (UTC)

OK, I didn't understand what you meant about 0s ... I perfectly agree, they are used only in some databases but never in scholarly articles, so we can probably avoid using them in aliases; using collection (P195) for the institution preserving the object and published in (P1433) for the edition(s) of the publication(s) containing its text sounds very good. --Epìdosis 15:55, 15 June 2022 (UTC)

@Epìdosis, Jahl de Vautban: could you please take a look to my draft of guidelines on the main page.--Alexmar983 (talk) 23:04, 20 June 2022 (UTC)

@Alexmar983: sorry, I have a limited access to a proper computer until the 30th, but with a rapid check your guidelines seem good enough to go ahead. —-Jahl de Vautban (talk) 19:44, 23 June 2022 (UTC)

More details in existing items:

1. in CIL III 007523 (Q102552060) culture (P2596) Ancient Rome (Q1747689) is also used;
2. in CIL VI 10281 (Q90151800) a qualifier for the language is directly in the P31 of the iscription.

Do you think it can be put in the general example?--Alexmar983 (talk) 21:43, 28 June 2022 (UTC)

@Alexmar983: 1) could be useful, but not vital ; 2) I don't see any value of having it as qualifier when we'll be using it anyway as a main property. --Jahl de Vautban (talk) 14:02, 2 July 2022 (UTC)

Qualifiers to instance of (P31) are widely discouraged; this is obviously redundant, existing as main property, thus I removed it - similar cases should go in the same way. Agree about culture (P2596)Ancient Rome (Q1747689) being useful but not fundamental. --Epìdosis 17:48, 2 July 2022 (UTC)

User:Epìdosis,User:Jahl de Vautban ok I can put it in the guideline.--Alexmar983 (talk) 18:05, 2 July 2022 (UTC)

Double code for same inscription on same catalogue

We can accept that there are two codes on different catalogues... on Commons we as a community used the CIL as a structured standard for Latin inscriptions for example, on Wikidata we will use multiple aliases and metadata. But what should we do if we have a double code for same inscription in the same catalogue, if two Commons categories are already created, see c:Category:CIL VI 000110 and c:Category:CIL_VI_030694.

This is not urgent but... should we go for a redirect? We can't have two wikidata items for the inscription. We can have one for the artifact, than separate one for the inscription, but the item for the inscription should be only one. So when we will have a item for the iscription different than the artifact itself, should we make a redirect on one category on Commons and link through one Wikidata Infobox to all aliases and metadata on just one item here?

For example I have created c:Category:CIL_V_000999 with Wikidata item, than discovered c:Category:CIL V 003997. Even if I separate the item from the inscription and the artifact, It's two commons category for one Wikidata item. --Alexmar983 (talk) 10:16, 1 July 2022 (UTC)

What about c:Category:AE 1991, 0843 and c:Category:CIL V 004948... in this casa it is even a different catalog, They both refer to the inscription, so again why having two categories if the Wikidata item can be only one as a concept and encode all metadata on the long term? Should we go for something more refined?--Alexmar983 (talk) 09:46, 6 July 2022 (UTC)

I ping User:Nonopoly, he might not be an expert on the details of Wikidata but he created the trees of categories years ago and it's time we also hear his opinion.--Alexmar983 (talk) 09:48, 6 July 2022 (UTC)

This is an important question, as it happens quite often. I think we need to distinguish between full duplicates and inscriptions published as separate parts, either because one of them was found only later or because epigraphists didn't knew they were about the same item. In the first cas, we can use redirects; in the second one, I'd keep both categories, for some items (like photos) may only refer to one of these part; I'd also create a new category (if possible c:CIL VI 000110 = 030694 for the whole thing, with the old catagories as sub categories. We can then rename it with a more fitting name in case both parts are republished as one inscription by an autoritative corpus. --Jahl de Vautban (talk) 07:47, 9 July 2022 (UTC)

Inscriptions

Do you prefer the style to be like in Q112815205 or like in Q112814070? Should we use capitol letters? Should the missing reconstructed part inserted between squared bracket or round brackets? Alexmar983 (talk) 10:45, 30 June 2022 (UTC)

Should we use it at all? Those aren't citation, but the full text of the inscription. I really wonder if the full text should be kept on Wikidata. Also, if we want to follow a rigorous edition based on the w:en:Leiden Conventions wen are going to run into issue with some character (like dotted letters). --Jahl de Vautban (talk) 14:04, 2 July 2022 (UTC)

Jahl de Vautban ok we should go for a standard probably without capital letters.--Alexmar983 (talk) 14:43, 2 July 2022 (UTC)

The stile of the text in quotation or excerpt (P7081) is a complex issue, I'm not sure to have enough knowledge of epigraphy to express a motivated opinion. Anyway, a good provisory standard should be: whichever use of P7081 should have a reference; the text is thus cited with the format it has in the reference. I would not deform the text given in a reference in order to adequate to a Wikidata standard of any sort and I would not add value without a reference. --Epìdosis 17:51, 2 July 2022 (UTC)

User:Epìdosis ok but 20% of images on Commons of CIL have a text so I will not get rid of such information and insert it even if not referenced; I will put it without reference and we will fix it later. However, I am fine with using this version in the final guideline.--Alexmar983 (talk) 18:04, 2 July 2022 (UTC)

Jahl de Vautban, User:Epìdosis Tomb No.97 Isola Sacra (Q112944867) use the unabbreviated text as qualifier. That could be a good standard in the model, right?--Alexmar983 (talk) 18:06, 6 July 2022 (UTC)

Also, I think we should switch to inscription (P1684) in the model and keep quotation or excerpt (P7081) for long inscriptions. So, we agree that the stadard for short text is only inscription (P1684). All CIL iscriptions could use that one.--Alexmar983 (talk) 18:15, 6 July 2022 (UTC)

OK for inscription (P1684), it makes more sense. For unabbreviated text (P7008), it could certainly be useful, but it should be with the original text, not the translation. --Jahl de Vautban (talk) 07:11, 9 July 2022 (UTC)

Part of...

Do you think I should create items for CIL volumes? I could insert proper topic in items such as Q21159086, and use P361 (part of) in every item of inscriptions to encode the volume label a metadata somehow. Alexmar983 (talk) 14:23, 30 June 2022 (UTC)

Yes we should! But I think published in (P1433) might be more fitting than part of (P361). --Jahl de Vautban (talk) 14:13, 2 July 2022 (UTC)

User talk:Jahl de Vautban so published in (P1433) and as argument the item of the volume?--Alexmar983 (talk) 14:45, 2 July 2022 (UTC)

@Alexmar983, Jahl de Vautban: using published in (P1433) with value the volume of CIL and qualifier section, verse, paragraph, or clause (P958) with value "VI, 1194" would be redundant, while using value "1194" would be somewhat ambiguous (1194 is the page or the number of the inscription?). I wounder if using published in (P1433)Corpus Inscriptionum Latinarum (Q691007) with qualifier section, verse, paragraph, or clause (P958) with value "CIL VI, 1194" or simply "VI, 1194" would maybe be a better solution; it avoids redundances or ambiguities and makes queries a bit easier, also not implying the immediate creation of the items for the volumes (which are, anyway, important to be created); the only problem of this last solution is that it doesn't create a structured link between the volume-item and the inscription-item; but it would be feasible to convert this solution into another one in the future by bot. What's your opinion? --Epìdosis 18:00, 2 July 2022 (UTC)

I can go with that.--Alexmar983 (talk) 18:16, 6 July 2022 (UTC)

@Epìdosis: I agree it would be redundant, but the problem you pointed is the reason I'd favor the redundancy over the absence of a link between volume and inscription. Now, I wonder if using section, verse, paragraph, or clause (P958) with only 1194 is as ambiguous as you said; the existence of page(s) (P304) makes me think that section, verse, paragraph, or clause (P958) is reserved for references that aren't pages. --Jahl de Vautban (talk) 07:18, 9 July 2022 (UTC)

Or perhaps, instead of section, verse, paragraph, or clause (P958) with published in (P1433), why not using catalog code (P528) with catalog (P972) as qualifier? (some thoughts based on recent changes on Q97345601) --Jahl de Vautban (talk) 07:23, 9 July 2022 (UTC)

Whatever is the final style we agree upon, should I in any case start to create items of the CIL volumes? Let me know.--Alexmar983 (talk) 08:44, 17 July 2022 (UTC)

Translated version with P7081

See for example Q112876133. Should the modern language included at the same "level" as the original one in your opinion? or some more sophisticated use of qualifier should be used? Alexmar983 (talk) 17:20, 3 July 2022 (UTC)

object has role (P3831)translated edition (Q39811647) as qualifier in my opinion. --Epìdosis 19:07, 7 July 2022 (UTC)

Dis Manibus

I think I will create an item from c:Category:Dis Manibus inscriptions, as an additional ibstance. Just informing you in advance. Alexmar983 (talk) 22:01, 6 July 2022 (UTC)

Would fit nicely as subclass of epitaph (Q1772). On it culture (P2596)Ancient Rome (Q1747689) would be useful. --Jahl de Vautban (talk) 07:29, 9 July 2022 (UTC)

 Done Dis Manibus inscription (Q113027699). Please decide how to link it to Q661713 as well. Not sure about the best possible P18. --Alexmar983 (talk) 15:01, 12 July 2022 (UTC)

P1801

Please consider the use of plaque image (P1801) in Virius Nicomachus Flavianus (Q966315). This is also the image of the object CIL VI, 1782 (Q112960273). Do we have a "commemorative plaque property"? I think it should be a more structured metadata and not a shared image. Alexmar983 (talk) 13:40, 7 July 2022 (UTC)

Is your purpose linking from Virius Nicomachus Flavianus (Q966315) to CIL VI, 1782 (Q112960273) through a property X or I misunderstood? --Epìdosis 19:10, 7 July 2022 (UTC)

User:Epìdosis yes, is there a P for that or the information is only left to the shared used of an image? is there a P1801 that should link to item and not Commons files?--Alexmar983 (talk) 19:13, 7 July 2022 (UTC)

Then I would qualify plaque image (P1801) with statement is subject of (P805) instead of creating a new property. But we can also hear other opinions. --Epìdosis 19:17, 7 July 2022 (UTC)

Fine with me, I will add it later to the guideline if nobody disagrees.--Alexmar983 (talk) 19:18, 7 July 2022 (UTC)

done.--Alexmar983 (talk) 18:57, 8 July 2022 (UTC)

New ID

I think the EDR database could generate a good stable ID here. Their webpage show structured information about discovery place, language text, abbreviation from other database. With such ID, we can improve the references.

In this section I have been listing all possible IDs related to this project, including some practical aspects. For example, some of them require additional search to get to the other codes of the text. This seems more elegant and complete. Alexmar983 (talk) 11:33, 13 July 2022 (UTC)

Based on linked databases, I can add :

• EDH
• Clauss-Slaby
• CIL

--Jahl de Vautban (talk) 16:52, 13 July 2022 (UTC)

And they have no property yet? I'll check later. --Alexmar983 (talk) 14:25, 16 July 2022 (UTC)

I will start to propose future properties from this list starting maybe from next week.--Alexmar983 (talk) 20:49, 17 July 2022 (UTC)

@Alexmar983, Jahl de Vautban: Has anything happened with this yet? It would be great to have links to the databases on every inscription item. --Tolanor (talk) 23:05, 16 December 2022 (UTC)

@Tolanor: so independently from this projet Epigraphic Database Heidelberg ID (P11273) and Packard Humanities Institute (PHI) Greek Inscriptions ID (P11299) were recently created. That should be good for now to create quite a few inscriptions in both Latin and Greek. --Jahl de Vautban (talk) 23:10, 16 December 2022 (UTC)

I have to inform on social media about some new IDs (I was preparing the infographics this afternoon with User:Epìdosis) but again, I will back around January. I plan to propose other IDs during 2023, but as a property creator I barely have time to remove backlog with current proposal. Beware, we will have a proper set of IDs but as far as I could check last summer, no database online is under CC-0 so it's a boring, long and manual work to use them, I see no future massive import at the moment. As soon as my current deadlines are over, I will come here again.--Alexmar983 (talk) 23:18, 16 December 2022 (UTC)

Other format for aliases of CIL

Hi, what do you think of this format here for CIL. Example: CIL 6.956 Alexmar983 (talk) 16:40, 13 July 2022 (UTC)

Made from material

I will add soon made from material (P186) to the bullet list of properties to consider. Suggesting marble (Q40861) and stone (Q22731) Alexmar983 (talk) 08:28, 15 July 2022 (UTC)

Good practice for abreviating corpora as labels and aliases

As corpora's abreviations can, were and are to some extant still a complicated mess, I suggest that this project aligns itself with AEIGL GrEpiAbbr recommendations. It was established by leading epigraphists, it is up-to-date, and it is fearly complete. --Jahl de Vautban (talk) 19:52, 24 July 2022 (UTC)

Fine we can put it in the guidelines.--Alexmar983 (talk) 23:05, 16 December 2022 (UTC)

Q110284871

Hi, is there an item for the inscription on this grave? If not, could someone create and link it? (So that I see how it's done.) Tolanor (talk) 23:04, 16 December 2022 (UTC)

Tolanor I will back full time on this project in January. Can you wait few weeks? In any case there is a stub of guidelines on the home page, I plan to add few more suggestions of some external IDs recently created, but it was proposed starting from CIL. As far as I can see from the category on Wikimedia Commons, this one is AE 1931, 0068AE 2006, 0931. More importantly, this is the problem of separating inscriptions from artifact, something that I am aware about. Again, if you can wait January, we can arrange some model xamples together.--Alexmar983 (talk) 23:10, 16 December 2022 (UTC)

Sure, I think the world can wait a couple of weeks for this new inscription item to be created ;-). Just ping me when you're ready. Happy holidays! --Tolanor (talk) 23:15, 16 December 2022 (UTC)